Automatic speech recognition (ASR) is the translation of spoken words into text. Speech recognition applications include voice user interfaces such as voice dialling, call routing, appliance control, search, data entry, preparation of structured documents, and speech-to-text processing (e.g., word processors or emails).
In some conventional applications, when using ASR to contact a desired recipient, often multiple contacts with the same name will be returned. Some of those contacts may not have the relevant contact fields populated for the intended type of communication. In some instances, a selection of one of those contact names may result in an error message or prompt which indicates that the selected contact does not have that type of contact field for the given application.
For example, in some conventional systems, if the user wishes to send an e-mail using only the first name of the intended recipient, all users having that name may be presented to the user, regardless of whether they have an associated e-mail address.
Some other conventional systems use complicated weighting and disambiguation algorithms, to attempt to locate the best result using ASR. Such systems can be prone to false positives, and rely on weighted probabilities to, often but not necessarily always, identify the intended recipient.
Additional difficulties with some existing systems may be appreciated in view of the detailed description below.
Like reference numerals are used throughout the Figures to denote similar elements and features.